Purification, characterization, and comparison of poliovirus RNA polymerase from native and recombinant sources

Abstract

Poliovirus RNA encodes an RNA-dependent RNA polymerase, designated 3Dpol, that catalyzes the synthesis of both plus and minus strand viral RNA. This enzyme was purified to near homogeneity from poliovirus-infected HeLa cells, recombinant baculovirus-infected insect cells, and from Escherichia coli transformed with an expression plasmid containing poliovirus 3D sequences. The two recombinant expression systems produced significantly higher yields of active enzyme than could be attained from virus-infected HeLa cells. All preparations contained a 52-kDa protein, recognized by antisera raised against 3D expressed as a fusion protein in E. coli. Immunoreactive protein resolved into 3-4 species on isoelectric focusing sodium dodecyl sulfate-polyacrylamide gel electrophoresis two-dimensional gels. Efforts to demonstrate that the multiple spots resulted from phosphorylation were negative. Furthermore, no evidence for autophosphorylation of purified 3Dpol was obtained. Purified 3Dpol from recombinant sources manifested the same specific activities as enzyme from poliovirus-infected HeLa cells in both a poly(A)-dependent poly(U) polymerase assay and a poliovirus RNA-dependent RNA polymerase assay. The products of the latter reaction reached the length of the template (7.5 kilobases) in 20-30 min, indicating an elongation rate of approximately 300 nucleotides/min at 30 degrees C. No products exceeded the length of the template. Intermediate length products were detected, which presumably resulted from pauses in transcription due to template structure. All transcription was dependent on primer. The kinetic parameters of all three purified enzyme preparations were the same; the Km for UTP was 2.4 +/- 0.1 microM in an RNA polymerase activity assay. Product formation was linear for up to 45 min, except for a 3-5-min lag before synthesis began. The lag was independent of enzyme concentration, and independent of the template used. The lag was eliminated by preincubating enzyme, template, primer, and three of the four nucleotide triphosphates, but not by preincubating any subset of these components. This suggested that a preinitiation complex must form as a prerequisite to RNA synthesis. Partially purified preparations of 3Dpol from the three sources showed significant differences in activities and products, including the appearance of primer-independent polymerase activity and production of dimer-length RNA products. These variable properties are likely due to different contaminating activities provided by the different cellular hosts, since upon further purification, all three enzymes exhibited identical properties.